Abstract: Through investigation and research on the monitoring section of the existing converter bag filter dust collection system in a steel plant, it was found that the system uses industrial fieldbus for communication, with multiple and inconsistent protocols, such as OPC, Modbus, and Modbus-Modbus Plus. The entire system is divided into seven parts, which are scattered and geographically distant, making effective system analysis and control difficult, resulting in low dust collection efficiency. To address this problem, Schneider Electric's Quantum series PLC and MP7 industrial monitoring software were adopted. Industrial Ethernet was used as the communication channel for field data, and TCP-IP was used as the unified communication protocol for centralized monitoring. Event activation was implemented using a trigger-based approach, thereby reducing system resource utilization and effectively minimizing network-induced latency. Through this system design, dust collection efficiency was improved, flue gas emission concentration was reduced, and atmospheric environmental quality was improved.
Keywords: bag filter; network monitoring; industrial Ethernet; redundancy; latency
0 Introduction
The large amounts of dust generated during steel smelting pose a significant threat to the health of operators and the environment. Baghouse dust collectors are widely used due to their effectiveness in dust control. Although a steel plant in Jinan uses a baghouse dust collector system, the lack of a comprehensive dust collector monitoring system results in fragmented and manual monitoring, leading to low dust collection efficiency, poor real-time system maintenance, heavy workload, and incomplete information. This also results in excessive personnel and wasted resources. Therefore, establishing a high-performance dust collector monitoring system is essential. Traditional methods using high-level languages such as VB or VC++ to develop monitoring systems offer advantages such as low-level development and high flexibility, but also suffer from long development cycles, poor system portability, and limited scalability. To address these shortcomings, this system utilizes MonitorPro V7.2 industrial control configuration software developed by Schneider Electric.
1. Overall System Network Topology
The dust removal network monitoring system for converters in groups 2 and 3 consists of 7 subsystems, including: auxiliary raw material dust removal system; blast furnace dust removal system; secondary dust removal system for converter group 1; first-phase LF dust removal system; hot metal pretreatment dust removal system; secondary dust removal systems for converters 2 and 3; and second-phase LF secondary dust removal system, as shown in Figure 1.
The seven subsystems are located in seven different sites, each consisting of a dust collector motor and dust collector body, a dust removal control system, a dust removal and ash unloading system, and a dust conveying system. The host computer and the field systems utilize a Schneider PLC control system (combining electrical and instrumentation functions) for sequential and continuous control, forming a Level I basic automation system. The Level II process computer consists of two server backup systems, whose main functions include process technology calculation, production guidance, process monitoring, data communication, and production reporting. An HMI (Human-Machine Interface) substation is set up near the site. The substation system communicates with the host computer via 10M/100M industrial Ethernet . Due to the relatively large distance between the field substations, a separate central monitoring room is required. By centralizing the seven subsystems in one monitoring room, data is collected and stored on a fixed server in a real-time SQL Server database. Visual Basic tools are used to automatically generate and print reports, thus automatically and in real-time recording of field operation, faults, curves, and personnel on-duty status.
Triggered by the rising edge of the voltage level, its output only determines the input at the edge moment, thus exhibiting strong anti-interference capability and preventing malfunctions. In this system, due to the complexity of the equipment and varying requirements, different triggering methods must be selected for different operational situations. For example, level triggering is sufficient for some digital quantity toggling operations; while for more critical operations such as starting and stopping equipment, pulse triggering is used to ensure safety and stability.
2. System Function Introduction
2.1 Equipment start-up and shutdown operations
There are three ways to start and stop the equipment: on-site, local, and centralized. Local and centralized operations are all performed using a mouse or keyboard. Centralized computer operation is used for equipment start and stop. Before starting, it's necessary to check if the local control cabinet has granted operating permissions; operation can only proceed if permissions are granted to prevent misoperation. In stand-alone or debugging mode, individual equipment can be operated, including automatic/manual operation selection, equipment start/stop control, and selection of dust cleaning and conveying methods.
2.2 Display and Alarm Prompts
The display section needs to show parameters such as shaft pressure, working oil pressure, vibration intensity, and stator temperature of the dust collector fan; the status of the solenoid valves, material level, and rapping in the dust collector chamber; and the pressure difference in the dust collector chamber. Different colors are used to distinguish the equipment's operating status, and process control parameters are simultaneously displayed in the corresponding positions. Multiple screens are designed to represent different production processes, with the mouse switching between the overall view and details of the process. In case of a fault, an audible and visual alarm is triggered, the alarm bar flashes to indicate the fault, and the fault location, time, and time are recorded in the background database for later review and analysis. After the fault is confirmed, the sound and flashing screen can be manually deactivated, but the screen remains active.
2.3 Parameter History
It records and generates historical curves and reports in real time, and saves data on important process parameters for production debugging, accident analysis, and querying.
2.4 System Security Management
To prevent unauthorized access to the computer and staff from logging out of the system, and to ensure the security of the entire monitoring system, corresponding user levels and passwords are designed for different users. This system employs a user management function with three levels: ordinary users, advanced users, and super users, each with different operating permissions. Users can only operate within the scope of the permissions granted by the super user, thereby enhancing the system's security and reliability.
2.5 Report Printing Function
To monitor production status and reduce operator workload, an automated reporting system was designed. Using Visual Basic, it reads data from a SQL database, automatically generates reports, and prints them. The printing methods include: ① Scheduled printing: Printing T/R operation reports at set intervals; ② User-defined printing: Users can customize the printing of current or historical reports; ③ Alarm printing: When a T/R malfunction occurs, the system automatically records the time of the malfunction, the type of malfunction, and the operating parameters at the time of the malfunction.
3 System Software Design
This system is mainly built on Schneider Electric's MonitorProV7-2 industrial control configuration software. The server end of the software and the lower-level PLC communicate via industrial Ethernet using the TCP/IP protocol, and the client end allows users to operate the system through a convenient human-machine interface.
3.1 Introduction to MonitorPro V7.2
MonitorPro V7.2 is an industrial configuration software developed by Schneider Electric, designed to optimize production management processes. It provides best-in-class technology for information management and processing in various fields—automation, agricultural sectors, water treatment, etc. Its modular design allows for customization to meet application requirements. Its user-friendly interface and flexible structure facilitate tailoring to individual user needs. Users can quickly create their own user interfaces, add remote maintenance capabilities, and perform additional operations.
3.2 Introduction to Industrial Ethernet
Ethernet is a local area network (LAN) introduced by Xerox PLC in 1975. Currently, because the Internet and Industrial Ethernet use the unified communication protocol—TCP/IP—Ethernet boasts advantages such as real-time performance, stability, and versatility. Therefore, it is increasingly used not only in the human-machine interface (HMI) layer and enterprise information system layer, but also in the I/O layer and HMI layer, earning the reputation of being "e-network to the end." Despite possessing the inherent drawbacks of all networks, such as issues with real-time and deterministic communication, robustness, and interference resistance, its advantages and disadvantages are comprehensively considered, leading to its widespread application in industrial settings.
3.3 Display Design for System Monitoring
Although the display part in the monitoring is relatively simple, in order to achieve a concise and clear picture, fewer tags, and easier management and maintenance, similar or identical devices are defined by arrays. Each type of device is replaced by a tag name. Furthermore, instead of defining each of the most numerous digital quantities in the system one by one, the method of defining multiple digital quantities by using each bit of the binary representation of the analog quantity to define the 0 and 1 changes of multiple digital quantities is not adopted, which greatly reduces the tedious tag definition process.
3.4 Design and Implementation of Redundant Servers
The on-site requirements necessitate continuous, fault-free system operation, necessitating a master server and a slave server (standby server) for redundancy. This requires two identical servers, configured to support master/slave arbitration, and synchronized real-time data, database records, and system alarms. On-site clients connect to both servers simultaneously. If either server fails, the system switches to the other within 10ms, ensuring uninterrupted real-time system operation and achieving fault-free, safe operation.
3.5 Solving Induced Delay in Networks
Due to the large number of points in the system, their long distances, and the inherent data latency issues of networks, system instability increases. Currently, common methods include collision fragmentation shielding and latency compensation. Considering that network-induced latency is caused by packet queuing delays, information generation delays, and transmission delays, a triggering method is used to request data transmission to reduce data waiting time, thereby reducing latency.
Generally, transmission uses two triggering methods: level triggering and pulse triggering. Different triggering methods have different advantages and disadvantages: ① Level triggering: Level triggering occurs when the signal becomes high or low, so its anti-interference capability is weak. It also has the advantages of simple structure and fast triggering speed. However, its disadvantage is that it is prone to causing a "ghosting" (or "skipping") effect. ② Pulse triggering: Pulse triggering occurs on the rising edge of the signal becoming high. Its output only determines the input at the edge, so it has strong anti-interference capability and does not cause a "ghosting" effect.
3.6 An Algorithm in Alarm Point Design
In the design of the system fault location and alarm, due to the involvement of valve linkages, a specialized algorithm was designed to ensure normal and accurate alarm operation. Taking the linkage and interlocking relationship between the dust collector's pulse valves and lifting valves as an example, the following explanation is provided: During dust cleaning, the lifting valve of the dust cleaning chamber is first closed, meaning the lifting valve of that chamber is energized. Then, the 12 pulse valves of that chamber are sequentially activated. Compressed air is injected into the filter bags through the pulse valves and blowpipes, causing the filter dust attached to the outside of the filter bags to fall off. After the 12 pulse valves of the dust cleaning chamber have finished blowing, the lifting valve of that chamber is opened, restoring the filtration state. Since the opening/closing of valves is a typical digital 0/1 switching, and there are many valves, based on the method of defining multiple digital quantities using each bit of an analog quantity's binary representation, one lifting valve in one chamber corresponds to 12 pulse valves and 12 nozzles, defined as an analog quantity me. The first to 12 bits of its binary representation represent the states of the 12 pulse valves, respectively.
The alarm rule is as follows: When both the dust collector lift valve and the pulse valve are working normally (i.e., both are displayed in green), the corresponding nozzle in the dust collector will also work normally (displayed in green, 1). If any one of them malfunctions (the dust collector lift valve or any pulse valve), the corresponding nozzle in the dust collector will not work normally (displayed in red, 0). To implement this rule, binary bit manipulation is used. `mc` represents the on/off state of the 12 pulse valves. Therefore, a bitwise AND operation is performed on the state bit of each pulse valve, and then a bitwise AND operation is performed with the state bit of the lift valve. When the lift valve state `ts` and the pulse valve state `me` are both present, the alarm will be triggered.
